pith. sign in

arxiv: 2604.07248 · v2 · submitted 2026-04-08 · ⚛️ physics.optics · cs.CV

TurPy: a physics-based and differentiable optical turbulence simulator for algorithmic development and system optimization

Joseph L. Greene , Alfred Moore , Iris Ochoa , Emily Kwan , Patrick Marano , Christopher R. Valenta This is my paper

Pith reviewed 2026-05-10 18:29 UTC · model grok-4.3

classification ⚛️ physics.optics cs.CV
keywords optical turbulencephase screendifferentiable simulatorscintillationGaussian beam broadeningfree-space opticswave opticsgradient optimization
0
0 comments X

The pith

TurPy reproduces Gaussian beam broadening and scintillation to 98 percent of closed-form models while enabling gradient-based optical system design.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

TurPy is presented as a GPU-accelerated wave optics simulator that generates phase screens for modeling turbulence distortions in free-space propagation. The code uses subharmonic generation, autoregressive evolution, and automated placement to produce realistic wavefront effects from basic medium properties. Full differentiability lets the simulator participate directly in gradient descent loops for training optical components or networks. Validation confirms second-order beam spreading and fourth-order intensity fluctuations match theory across weak to strong regimes. This setup supports synthetic data creation and end-to-end optimization for platforms that must operate through turbulent paths.

Core claim

TurPy is a physics-based and differentiable optical turbulence simulator that incorporates subharmonic phase screen generation, autoregressive temporal evolution, and an automated screen placement routine. Parameterized solely by the medium's refractive index structure constant and power spectral density, it extends to various propagation environments. It matches second-order Gaussian beam broadening and fourth-order plane wave scintillation to closed-form models with 98 percent accuracy from weak to strong turbulence. As a demonstration, TurPy optimizes a dual-domain diffractive deep neural network to recover a Gaussian beam from a weakly turbulent path, achieving over 20 times reduction in

What carries the argument

Subharmonic phase screen generation parameterized by power spectral density, combined with automated placement and full differentiability for end-to-end gradient optimization.

If this is right

  • Optical systems for free-space links can be trained end-to-end inside the simulator to compensate for turbulence without separate calibration steps.
  • Changing only the input power spectral density allows the same code to model propagation in oceanic or biological media.
  • High-fidelity synthetic datasets become available for developing and testing turbulence-aware image or signal processing algorithms.
  • Time-correlated turbulence sequences can be generated for applications that require dynamic rather than static distortion modeling.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The framework could be coupled to hardware-in-the-loop testbeds to refine designs before physical deployment.
  • Similar parameterization might extend the approach to acoustic or radio-wave propagation through turbulent media.
  • It opens the possibility of learning adaptive optics corrections directly from gradient signals produced by the simulator.
  • Satellite-to-ground or drone-based optical systems could use the tool to explore trade-offs in aperture size and correction speed.

Load-bearing premise

The subharmonic phase screen generation and automated placement routine correctly balance Fourier aliasing and weak-turbulence approximations for the chosen propagation distances and turbulence strengths.

What would settle it

A side-by-side comparison of TurPy-simulated scintillation values against independent laboratory measurements in a controlled turbulence chamber that deviates by more than two percent from the reported match to theory.

Figures

Figures reproduced from arXiv: 2604.07248 by Alfred Moore, Christopher R. Valenta, Emily Kwan, Iris Ochoa, Joseph L. Greene, Patrick Marano.

Figure 1
Figure 1. Figure 1: Overview of Phase Screen Generation. (A) [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. Figure 3: Autoregression for Phase Screen Evolution. (A) [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Recovery of Turbulence PSD. (A) Input PSD. (B) MSE between input and recovery PSD. (C) Example recovered PSD after TurPy phase screen averaging. 3.2 2 nd Order Statistics Validation through Gaussian Beam Broadening 2nd order statistics describe key turbulence-induced beam properties including broadening, wander, angle of arrival, and spatial coherence loss. We analyze Gaussian beam broadening due to its pr… view at source ↗
Figure 5
Figure 5. Figure 5: 2nd Order Broadening of a Gaussian Beam. (A) TurPy simulated free space and turbulent Gaussian beam size (e.g., waist) versus theory for weak turbulence, (B) moderate turbulence, and (C) strong turbulence over paths of increasing length. (D) Example averaged TurPy outputs to recover an averaged broadened Gaussian for characterization. 3.3 4 th Order Statistics Validation through Plane Wave Scintillation 4t… view at source ↗
Figure 6
Figure 6. Figure 6: 4th Order Scintillation of a Plane Wave. (A) Average scintillation index from a plane wave passing through a turbulent path with variable strength. (B) Example of underdeveloped caustics generated in the weak turbulence regime. (C) Example of well-developed speckles generated in the strong turbulence regime. 4. TRAINING DIFFRACTIVE DEEP NEURAL NETWORKS FOR ALL-PASSIVE ADAPTIVE OPTICS ALTERNATIVES To demons… view at source ↗
Figure 7
Figure 7. Figure 7: Training of a D2NN for Beam Stabilization over Turbulence. (A) [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
read the original abstract

Developing optical systems for free-space applications requires simulation tools that accurately capture turbulence-induced wavefront distortions and support gradient-based optimization. Here we introduce TurPy, a GPU-accelerated, fully differentiable wave optics turbulence simulator to bridge high fidelity simulation with end-to-end optical system design. TurPy incorporates subharmonic phase screen generation, autoregressive temporal evolution, and an automated screen placement routine balancing Fourier aliasing constraints and weak-turbulence approximations into a unified, user-ready framework. Because TurPy's phase screen generation is parameterized through a media-specific power spectral density, the framework extends to atmospheric, oceanic, and biological propagation environments with minimal modification. We validate TurPy against established atmospheric turbulence theory by matching 2nd order Gaussian beam broadening and 4th order plane wave scintillation to closed-form models with 98% accuracy across weak to strong turbulence regimes, requiring only the medium's refractive index structure constant and power spectral density as inputs. To demonstrate TurPy as a gradient-based training platform, we optimize a dual-domain diffractive deep neural network (D2NN) in a two-mask dual-domain architecture to recover a Gaussian beam from a weakly turbulent path and achieving over 20x reduction in scintillation relative to an uncompensated receiver in simulation. TurPy is released as an open-source package to support synthetic data generation, turbulence-informed algorithm development, and the end-to-end design of optical platforms operating in turbulent environments.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The paper introduces TurPy, a GPU-accelerated and fully differentiable wave-optics turbulence simulator built on subharmonic phase-screen generation, autoregressive temporal evolution, and an automated screen-placement routine that balances Fourier aliasing against weak-turbulence approximations. It claims 98% quantitative agreement with closed-form expressions for second-order Gaussian-beam broadening and fourth-order plane-wave scintillation across weak-to-strong regimes when supplied only with the refractive-index structure constant C_n^2 and the medium power spectral density. The work further demonstrates end-to-end gradient-based optimization of a two-mask dual-domain diffractive neural network that recovers a Gaussian beam and yields >20× scintillation reduction relative to an uncompensated receiver.

Significance. If the validation is robust, TurPy supplies a genuinely useful open-source platform that couples high-fidelity physics simulation with differentiable programming, enabling turbulence-aware algorithmic design and system optimization. The parameterization by external C_n^2 and PSD (rather than internally fitted quantities) and the explicit release as an open-source package are concrete strengths that support reproducibility and extension to non-atmospheric media.

major comments (1)
  1. [Validation and methods] Validation section (and the automated placement routine described in the methods): the headline 98% accuracy claim spans weak-to-strong turbulence, yet the screen-placement logic invokes a weak-turbulence approximation to set inter-screen distances. No aliasing-error budget, convergence study versus number of screens, or comparison against an independent reference propagator that avoids the same approximation is reported for the strong-turbulence cases. This leaves open the possibility that the quoted accuracy is sensitive to the particular distances and C_n^2 values chosen rather than generally robust.
minor comments (2)
  1. [Abstract] The abstract states 'over 20x reduction in scintillation' without specifying the exact metric (e.g., variance, peak intensity, or Strehl) or the precise baseline receiver; a short clarifying sentence would improve precision.
  2. [Figures] Figure captions and axis labels for the validation plots should explicitly state the number of Monte-Carlo realizations and the turbulence-strength parameter range (e.g., Rytov variance) used to generate the 98% figure.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their constructive and detailed review. We address the single major comment below and have revised the manuscript to incorporate additional validation material.

read point-by-point responses

  1. Referee: [Validation and methods] Validation section (and the automated placement routine described in the methods): the headline 98% accuracy claim spans weak-to-strong turbulence, yet the screen-placement logic invokes a weak-turbulence approximation to set inter-screen distances. No aliasing-error budget, convergence study versus number of screens, or comparison against an independent reference propagator that avoids the same approximation is reported for the strong-turbulence cases. This leaves open the possibility that the quoted accuracy is sensitive to the particular distances and C_n^2 values chosen rather than generally robust.

    Authors: The automated screen-placement routine selects inter-screen distances so that the weak-turbulence approximation holds locally for each segment; this is a standard and physically justified approach in split-step Fourier propagation, allowing the cumulative effect of many segments to reproduce strong turbulence. The 98% quantitative match to closed-form expressions for Gaussian-beam broadening and plane-wave scintillation is obtained against independent theoretical models that do not rely on the same placement approximation, providing evidence that the reported accuracy is not an artifact of the chosen distances or C_n^2 values. Nevertheless, to directly respond to the concern, the revised manuscript now includes an explicit aliasing-error budget and a convergence study versus number of screens for representative strong-turbulence cases. These additions confirm robustness across the tested parameter space. A comparison against a completely independent reference propagator that avoids any weak-turbulence element would require an entirely separate numerical framework and is therefore not provided here. revision: yes

Circularity Check

0 steps flagged

No significant circularity: validation uses independent closed-form theory

full rationale

The paper introduces a physics-based simulator whose core validation compares outputs (Gaussian beam broadening and plane-wave scintillation) to established closed-form atmospheric turbulence models, using only external inputs (refractive index structure constant C_n^2 and power spectral density). This matching is presented as an empirical check rather than a definitional or fitted result. No load-bearing steps reduce by construction to self-inputs, self-citations, or ansatzes; the automated placement routine is an implementation choice whose accuracy is asserted via comparison to external theory, not enforced by the claim itself. The derivation chain remains self-contained against independent benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The central claim rests on standard wave-optics propagation assumptions and the accuracy of the chosen power spectral density model; no new physical constants or entities are introduced beyond the input turbulence parameters.

free parameters (1)
  • refractive index structure constant C_n^2
    User-provided input that sets turbulence strength; not fitted inside the simulator.
axioms (2)
  • domain assumption Phase screens generated from the medium's power spectral density accurately represent the statistical properties of turbulence.
    Invoked in the validation section to justify matching closed-form models.
  • domain assumption Weak-turbulence approximations remain valid for the chosen screen spacing and propagation distances.
    Used in the automated screen placement routine.

pith-pipeline@v0.9.0 · 5571 in / 1396 out tokens · 38400 ms · 2026-05-10T18:29:49.420530+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

33 extracted references · 33 canonical work pages

  1. [1]

    L. C. Andrews and M. K. Beason, Laser beam propagation in random media: new and advanced topics. 2023. doi: 10.1117/3.2643989

    work page doi:10.1117/3.2643989 2023

  2. [2]

    Light Propagation in a Turbulent Ocean,

    “Light Propagation in a Turbulent Ocean,” in Progress in Optics, vol. 64, Elsevier, 2019, pp. 1–43. doi: 10.1016/bs.po.2018.09.001

    work page doi:10.1016/bs.po.2018.09.001 2019

  3. [3]

    Turbulent nature of refractive-index variations in biological tissue,

    J. M. Schmitt and G. Kumar, “Turbulent nature of refractive-index variations in biological tissue,” Opt. Lett., vol. 21, no. 16, pp. 1310–1312, Aug. 1996, doi: 10.1364/OL.21.001310

    work page doi:10.1364/ol.21.001310 1996

  4. [4]

    Laser Beam Atmospheric Propagation Modelling for Aerospace LIDAR Applications,

    T. Fahey, M. Islam, A. Gardi, and R. Sabatini, “Laser Beam Atmospheric Propagation Modelling for Aerospace LIDAR Applications,” Atmosphere, vol. 12, no. 7, p. 918, Jul. 2021, doi: 10.3390/atmos12070918

    work page doi:10.3390/atmos12070918 2021

  5. [5]

    High-power lasers for directed-energy applications,

    P. Sprangle, B. Hafizi, A. Ting, and R. Fischer, “High-power lasers for directed-energy applications,” Appl. Opt., vol. 54, no. 31, pp. F201–F209, Nov. 2015, doi: 10.1364/AO.54.00F201

    work page doi:10.1364/ao.54.00f201 2015

  6. [6]

    Optical communications in turbulence: a tutorial,

    L. B. Stotts and L. C. Andrews, “Optical communications in turbulence: a tutorial,” Opt. Eng., vol. 63, no. 4, p. 041207, Dec. 2023, doi: 10.1117/1.OE.63.4.041207

    work page doi:10.1117/1.oe.63.4.041207 2023

  7. [7]

    Structured light in atmospheric turbulence—a guide to its digital implementation: tutorial,

    C. Peters, V. Cocotos, and A. Forbes, “Structured light in atmospheric turbulence—a guide to its digital implementation: tutorial,” Adv. Opt. Photonics, vol. 17, no. 1, pp. 113–184, Mar. 2025, doi: 10.1364/AOP.538883

    work page doi:10.1364/aop.538883 2025

  8. [8]

    Simulation of a Kolmogorov phase screen,

    R. G. Lane, A. Glindemann, and J. C. Dainty, “Simulation of a Kolmogorov phase screen,” Waves Random Media, vol. 2, no. 3, pp. 209–224, Jul. 1992, doi: 10.1088/0959-7174/2/3/003

    work page doi:10.1088/0959-7174/2/3/003 1992

  9. [9]

    Log-amplitude variance for a Gaussian-beam wave propagating through non-Kolmogorov turbulence,

    L. Tan, W. Du, J. Ma, S. Y. U, and Q. Han, “Log-amplitude variance for a Gaussian-beam wave propagating through non-Kolmogorov turbulence,” Opt. Express, vol. 18, no. 2, pp. 451–462, Jan. 2010, doi: 10.1364/OE.18.000451

    work page doi:10.1364/oe.18.000451 2010

  10. [10]

    Computationally efficient autoregressive method for generating phase screens with frozen flow and turbulence in optical simulations,

    S. Srinath, L. A. Poyneer, A. R. Rudy, and S. M. Ammons, “Computationally efficient autoregressive method for generating phase screens with frozen flow and turbulence in optical simulations,” Opt. Express, vol. 23, no. 26, pp. 33335–33349, Dec. 2015, doi: 10.1364/OE.23.033335

    work page doi:10.1364/oe.23.033335 2015

  11. [11]

    Method for simulating atmospheric turbulence phase effects for multiple time slices and anisoplanatic conditions,

    M. C. Roggemann, B. M. Welsh, D. Montera, and T. A. Rhoadarmer, “Method for simulating atmospheric turbulence phase effects for multiple time slices and anisoplanatic conditions,” Appl. Opt., vol. 34, no. 20, pp. 4037–4051, Jul. 1995, doi: 10.1364/AO.34.004037

    work page doi:10.1364/ao.34.004037 1995

  12. [12]

    Partially coherent beam propagation in atmospheric turbulence [Invited],

    G. Gbur, “Partially coherent beam propagation in atmospheric turbulence [Invited],” JOSA A, vol. 31, no. 9, pp. 2038–2045, Sep. 2014, doi: 10.1364/JOSAA.31.002038

    work page doi:10.1364/josaa.31.002038 2038

  13. [13]

    Comparison of turbulence mitigation algorithms,

    S. T. Kozacik, A. L. Paolini, A. Sherman, J. L. Bonnett, and E. J. Kelmelis, “Comparison of turbulence mitigation algorithms,” Opt. Eng., vol. 56, no. 7, p. 071507, Mar. 2017, doi: 10.1117/1.OE.56.7.071507

    work page doi:10.1117/1.oe.56.7.071507 2017

  14. [14]

    Performance analysis of an adaptive optics system for free-space optics communication through atmospheric turbulence,

    Y. Wang et al., “Performance analysis of an adaptive optics system for free-space optics communication through atmospheric turbulence,” Sci. Rep., vol. 8, no. 1, p. 1124, Jan. 2018, doi: 10.1038/s41598-018-19559-9

    work page doi:10.1038/s41598-018-19559-9 2018

  15. [15]

    On the use of deep learning for computational imaging,

    G. Barbastathis, A. Ozcan, and G. Situ, “On the use of deep learning for computational imaging,” Optica, vol. 6, no. 8, pp. 921–943, Aug. 2019, doi: 10.1364/OPTICA.6.000921

    work page doi:10.1364/optica.6.000921 2019

  16. [16]

    A Differentiable Wave Optics Model for End-to-End Computational Imaging System Optimization,

    C.-J. Ho, Y. Belhe, S. Rotenberg, R. Ramamoorthi, T.-M. Li, and N. Antipa, “A Differentiable Wave Optics Model for End-to-End Computational Imaging System Optimization,” presented at the Proceedings of the IEEE/CVF International Conference on Computer Vision, 2025, pp. 28042–28051. Accessed: Feb. 03, 2026. [Online]. Available: https://openaccess.thecvf.co...

    work page 2025

  17. [17]

    A PyTorch-enabled tool for synthetic event camera data generation and algorithm development,

    J. L. Greene, A. Kar, I. Galindo, E. Quiles, E. Chen, and M. Anderson, “A PyTorch-enabled tool for synthetic event camera data generation and algorithm development,” in Synthetic Data for Artificial Intelligence and Machine Learning: Tools, Techniques, and Applications III, SPIE, May 2025, pp. 123–143. doi: 10.1117/12.3053238

    work page doi:10.1117/12.3053238 2025

  18. [18]

    Metasurface-Based Image Classification Using Diffractive Deep Neural Network,

    K. Cheng et al., “Metasurface-Based Image Classification Using Diffractive Deep Neural Network,” Nanomaterials, vol. 14, no. 22, Nov. 2024, doi: 10.3390/nano14221812

    work page doi:10.3390/nano14221812 2024

  19. [19]

    Multiple-image encryption and hiding with an optical diffractive neural network,

    Y. Gao, S. Jiao, J. Fang, T. Lei, Z. Xie, and X. Yuan, “Multiple-image encryption and hiding with an optical diffractive neural network,” Opt. Commun., vol. 463, p. 125476, May 2020, doi: 10.1016/j.optcom.2020.125476

    work page doi:10.1016/j.optcom.2020.125476 2020

  20. [20]

    Misalignment resilient diffractive optical networks,

    D. Mengu, Y. Zhao, N. T. Yardimci, Y. Rivenson, M. Jarrahi, and A. Ozcan, “Misalignment resilient diffractive optical networks,” Nanophotonics, vol. 9, no. 13, pp. 4207–4219, Oct. 2020, doi: 10.1515/nanoph-2020-0291

    work page doi:10.1515/nanoph-2020-0291 2020

  21. [21]

    On the general equivalence of the Fried parameter and coherence radius for non- Kolmogorov and oceanic turbulence,

    I. Toselli and S. Gladysz, “On the general equivalence of the Fried parameter and coherence radius for non- Kolmogorov and oceanic turbulence,” OSA Contin., vol. 2, no. 1, pp. 43–48, Jan. 2019, doi: 10.1364/OSAC.2.000043

    work page doi:10.1364/osac.2.000043 2019

  22. [22]

    Fast Approximations of Shift-Variant Blur,

    L. Denis, E. Thiébaut, F. Soulez, J.-M. Becker, and R. Mourya, “Fast Approximations of Shift-Variant Blur,” Int. J. Comput. Vis., vol. 115, no. 3, pp. 253–278, Dec. 2015, doi: 10.1007/s11263-015-0817-x

    work page doi:10.1007/s11263-015-0817-x 2015

  23. [23]

    AOtools: a Python package for adaptive optics modelling and analysis,

    M. J. Townson, O. J. D. Farley, G. O. de Xivry, J. Osborn, and A. P. Reeves, “AOtools: a Python package for adaptive optics modelling and analysis,” Opt. Express, vol. 27, no. 22, pp. 31316–31329, Oct. 2019, doi: 10.1364/OE.27.031316

    work page doi:10.1364/oe.27.031316 2019

  24. [24]

    Adaptive Optics Systems VI , year = 2018, editor =

    E. H. Por, S. Y. Haffert, V. M. Radhakrishnan, D. S. Doelman, M. van Kooten, and S. P. Bos, “High Contrast Imaging for Python (HCIPy): an open-source adaptive optics and coronagraph simulator,” in Adaptive Optics Systems VI, SPIE, Jul. 2018, pp. 1112–1125. doi: 10.1117/12.2314407

    work page doi:10.1117/12.2314407 2018

  25. [25]

    WavePy: a Python package for wave optics,

    J. Beck, C. Bekins, and J. P. Bos, “WavePy: a Python package for wave optics,” in Long-Range Imaging, SPIE, May 2016, pp. 8–17. doi: 10.1117/12.2225769

    work page doi:10.1117/12.2225769 2016

  26. [26]

    Wave optics simulator for lasers in time-evolving turbulence,

    G. Badura, C. Fernandez, and J. Stewart, “Wave optics simulator for lasers in time-evolving turbulence,” in Laser Radar Technology and Applications XXV, SPIE, Apr. 2020, pp. 18–29. doi: 10.1117/12.2558922

    work page doi:10.1117/12.2558922 2020

  27. [27]

    POPPY: Physical Optics Propagation in PYthon,

    M. Perrin, J. Long, E. Douglas, A. Sivaramakrishnan, C. Slocum, and others, “POPPY: Physical Optics Propagation in PYthon,” Astrophys. Source Code Libr., p. ascl:1602.018, Feb. 2016

    work page 2016

  28. [28]

    Diffractio: an open-source library for diffraction and interference calculations,

    L. M. Sanchez-Brea et al., “Diffractio: an open-source library for diffraction and interference calculations,” in Optics and Photonics for Advanced Dimensional Metrology III, SPIE, Jun. 2024, pp. 236–243. doi: 10.1117/12.3021879

    work page doi:10.1117/12.3021879 2024

  29. [29]

    TorchOptics: An open-source Python library for differentiable Fourier optics simulations,

    M. J. Filipovich and A. I. Lvovsky, “TorchOptics: An open-source Python library for differentiable Fourier optics simulations,” Nov. 27, 2024, arXiv: arXiv:2411.18591. doi: 10.48550/arXiv.2411.18591

    work page doi:10.48550/arxiv.2411.18591 2024

  30. [30]

    Baek, shwbaek/pado

    S.-H. Baek, shwbaek/pado. (Mar. 19, 2026). Python. Accessed: Apr. 03, 2026. [Online]. Available: https://github.com/shwbaek/pado

    work page 2026

  31. [31]

    Inner–outer interactions of large-scale structures in turbulent channel flow,

    J. Hwang, J. Lee, H. J. Sung, and T. A. Zaki, “Inner–outer interactions of large-scale structures in turbulent channel flow,” J. Fluid Mech., vol. 790, pp. 128–157, Mar. 2016, doi: 10.1017/jfm.2016.3

    work page doi:10.1017/jfm.2016.3 2016

  32. [32]

    Nonlinear Dependence of the Phase Screen Structure Function on the Atmospheric Layer Thickness,

    R. J. Mathar, “Nonlinear Dependence of the Phase Screen Structure Function on the Atmospheric Layer Thickness,” Feb. 27, 2011, arXiv: arXiv:1102.5513. doi: 10.48550/arXiv.1102.5513

    work page doi:10.48550/arxiv.1102.5513 2011

  33. [33]

    Non-Classic Atmospheric Optical Turbulence: Review,

    O. Korotkova and I. Toselli, “Non-Classic Atmospheric Optical Turbulence: Review,” Appl. Sci., vol. 11, no. 18, p. 8487, Jan. 2021, doi: 10.3390/app11188487. Appendix A: Comparison of TurPy to Other Simulation Engines Here, TurPy is compared to other python-based open-source wave optics simulation engines adaptable to turbulence to ground its functionalit...

    work page doi:10.3390/app11188487 2021